Observed Signal · Sep 16, 2026 · Product Launch · Source: AINews swyx · Impact: 4/5 · Sentiment: Positive
TypeSafe Launches Jev, a Fast Decision Model for AI Systems
TypeSafe AI, founded by former OpenAI researcher Diogo Almeida, launched Jev, a 'System One' model for fast, structured decision-making. Jev returns typed probabilistic outputs with confidence scores for predefined choices, avoiding text generation and thus eliminating token-by-token decoding, hallucinations, and the need for validators. Trained with a novel RLCD technique, it offers well-calibrated, confident decisions. Jev is 20-200x faster and 40-400x cheaper than frontier LLMs, with response times of 70-500ms, free output tokens, and input costs of $0.042/M. The launch gained massive attention, but integrations with Vercel, Cloudflare, LangChain, and others have solidified developer interest. TypeSafe raised $40M seed funding led by DCVC, valuing it at ~$200M. Now available without a waitlist, Jev is used for ad analysis, agent reasoning, and on-chain trading, with Vercel AI Gateway offering free access until September 25.
This is a major technical release from a prominent AI researcher (Diogo Almeida), introducing a new category of decision models that could significantly impact AI infrastructure and cost-efficiency in production systems.
Track TypeSafe AI Signals & Market Shifts in Real-Time
Polaris7 autonomous intelligence agents track regulatory filings, primary sources, executive changes, and deal flow 24/7. Create your free Explorer workspace to monitor these entities.
Key Takeaways & Evidence Grounding
- TypeSafe AI, founded by ex-OpenAI researcher Diogo Almeida, launched Jev, a decision-making model, out of stealth; it was released on September 15, 2026.
- Jev does not generate text; it returns typed probabilistic outputs with confidence scores for predefined choices, trained using RLCD (reinforcement learning for calibrated decisions), and claims zero hallucinations.
- Jev is 20-200x faster and 40-400x cheaper than frontier LLMs (up to 194x faster in evals, and in a benchmark returned 27 typed answers in 0.114s for $0.000081), with 70-500ms response times, free output tokens, and input costs of $0.042 per million tokens.
- Jev is integrated with Vercel, Cloudflare, LangChain, Langfuse, and OpenRouter; Vercel AI Gateway is offering free unlimited access until September 25, and it is now available to everyone without a waitlist.
- TypeSafe raised $40M in seed funding led by DCVC (valuation ~$200M), and Jev has seen applications in ad analysis, agent reasoning, classification, routing, scoring, verification, browser automation, and on-chain trading.
Connected Companies & Entities
29 Entities mapped“TypeSafe’s launch has sat comfortably atop Hacker News all day....”
“Google launched Gemini 3.8 Live and 3.8 Live Extended Thinking....”
“LangChain announced that every Managed Deep Agent is now an MCP server....”
“…frontier models like GPT-6 Astra are referenced in the context of Periodic Labs' Neon surpassing them....”
“Perplexity says it built and deployed CobbleDB, a DynamoDB replacement....”
“A notable Microsoft paper summary argues that on agent benchmarks, bash alone outperformed typed tool catalogs....”
“Meta’s argument that labs should invest heavily in alignment and external evaluation....”
Ontology Mapping & Concepts
Related Market Signals & Shifts
Recent verified developments and strategic activity across this market segment.
OpenAI's Decisions API Mirrors Jev, Sparks Competition in Agent Monitoring
At OpenAI's Dev Day, CEO Sam Altman announced the Decisions API, a limited preview product that provides similar functionality to TypeSafe AI's Jev, a model designed for software automation that outputs predefined choices as probabilities. The API aims to make model choices extremely fast while preserving capabilities. TypeSafe's CEO Diogo Almeida joked about 'clone wars' and sees OpenAI's interest as validation of its System One approach. The article highlights the emerging market for such decision models, which are cheaper and faster than frontier LLMs, and their potential application in monitoring and securing AI agents. A demo by QueryStory showed Jev could monitor agentic actions at a fraction of the cost of frontier LLMs, potentially preventing incidents like the Hugging Face one. OpenAI's security measures now include separate models to watch for bad actions, and Jev-like models could make this much more economical.
GPTBots.ai Integrates TypeSafe AI's Jev Decision Model
Aurora Mobile's enterprise AI agent platform, GPTBots.ai, has integrated Jev, a 'System One' decision model from TypeSafe AI, creating a two-layer AI architecture that separates reasoning from decision-making. Jev handles high-volume judgment tasks such as routing, filtering, and classification, returning structured probabilistic decisions in under 500ms at a fraction of the cost of a full LLM call. This integration powers three existing GPTBots.ai capabilities: Model Auto-Router, Dynamic Top-K for RAG, and Intent Classification in FlowAgent and Workflow. The move follows Jev's launch on September 15, 2026, and its rapid adoption by platforms like Vercel, Cloudflare, and LangChain. The integration aims to reduce cost, lower latency, and provide calibrated confidence scores for enterprise AI workflows, enabling more efficient and reliable automation.
Jev Decision Model: Fast, Cheap Classification for AI Pipelines
Claire Vo, founder of ChatPRD, demonstrates the new Jev decision model from TypeSafe AI in a video walkthrough. Unlike standard LLMs that generate text, Jev returns type-safe structured values (choice, score, probability) at a cost of $0.04 per million input tokens with no output charge. Vo details five real-world applications: categorizing 1,700 PRs for $0.09, analyzing local Claude Code and Codex sessions, triaging Gmail, building a product insights graph from 1,100 signals with 200,000 classifications, and creating a live audience dashboard from 4,500 YouTube comments. She emphasizes combining Jev's fast, cheap classification with more capable models like GPT-6 Astra for analysis and generation, achieving cost-effective and performant AI workflows. The video also demonstrates a real-time voice-to-color emotion mapping app, highlighting Jev's low latency for interactive use cases.
Track Real-Time Market Signals & Shifts
Set up custom watchlists to receive automated, evidence-grounded executive digests whenever material signals or shifts occur across your tracked landscape.